Why robots must be human-like to increase customer usage

Customers prefer robots to have human-like characteristics when dealing with them in customer service settings. They prefer robots to have a human voice, show emotions, and physical embody a human.

Innovative companies are always looking at new ways to make their customer interactions more efficient, engaging for the customer, but also cost-effective too. And the ever-increasing development of new technologies such as AI, machine learning and robotics have all allowed for companies to heavily invest in these practices, heavily impacting their customer interactions.


Take online chat bots for instance. Whether it be in a company’s online customer services or a financial and banking advice role, chatbots have been pretty widely adopted online to help consumers out, as the technology is more cost-effective than hiring someone for the role, and allows more customers to be helped out at once, using a series of algorithms from questions answered by the consumer.


But not all robots are as widely accepted or adopted. In Japan, the Henn Na hotel is one where all customer service staff from the receptionist at check-in to the bellboy are robots. Though it’s an incredible novelty and attracts a wide number of customers from across the world, the likelihood of this becoming the norm is unclear. It’s maybe be that we’ll keep robots on customer service desks and that’s all.


This is, of course, because using robots in customer service may not be everyone’s cup of tea, meaning that they may not have enough uptake or acceptance from the wider public to become the accepted norm. Often there is push back on the wide usage of new technologies, firstly because it could cause some employees roles to become redundant, but secondly because customers may not want to adopt the technology.


In fact, uptake may be low if the robots are difficult to use, if a huge number of customers already have negative attitudes towards robots, or lack the competence with technology and suffer technology anxiety. Effectively, if your main target customers are an older audience, perhaps it may not be a good idea to completely roll out robotised customer service.


But for many companies the strive to be as innovative as possible in their customer services interactions, and the want to implement new cost-effective, easy to use and efficient robotisation will always be there. So how can companies ensure that their customer service robots are not only good for the business but are widely accepted and adopted by their customers?


Alongside co-researchers from International Business School Suzhou, Paderborn University and the University of Rostock, we used a dataset of 11,053 individuals interacting with service robots from previous studies to find out the most effective way to design a customer service robot for maximum uptake.


The main link we wanted to investigate was the relationship between anthropomorphism – the attribution of human characteristics or behaviour to an object – and its consequences on robot perception and usage intention. By doing so, we could understand exactly what customers wanted from robot service interactions, what they didn’t want, and also the differences in their wants determined by their customer traits, such as technology anxiety and competence, sociodemographics (e.g., gender, age etc.), and the robot’s design features (e.g., physical, nonphysical) on anthropomorphism perception.


From the dataset, we found that customers were far more receptive to using a customer services robot when it had human-like characteristics such as a human voice, showed actual human emotions, and when it physically embodied a human not a robot. This is because customers find it easier to interact with robots who appear human-like as they are able to apply the familiar social rules and expectations of human–human interactions.


When a robot is perceived to be human-like it can better to ease and facilitate human–robot interactions. During a human–robot interaction where the robot is human-like, people can easily apply the social scripts and expectations of a human– human interaction, thus they tend to find the robot more controllable and predictable, and the interaction easier and more familiar. If people feel as though they are comfortable and at ease with a robot, their chances of using the service increase.


Many companies have been keen to introduce further robotisation into their practices, and these findings show that the perception of humanlike qualities in service robots can facilitate engagement with customers, as it incorporates the underlying principles and expectations people use in social settings in a person’s interaction with social robots.


For firms intending to employ service robots on the front line of their customer interactions, they should really consider employing humanlike versus machinelike robots. This will ensure maximum uptake and usage of these robots, and ensure that they are not a complete waste of time and effort.



Markus Blut is a Professor in Marketing at Durham Business School.His research interests are related to service marketing, retail management, and international business. He is particularly interested in new service technologies, international service marketing, retailing strategies, online retailing, consumer behavior, and relationship marketing.